WALDO

Welcome to the official website of Waldo, a cutting-edge open-source tool for automated detection of Adverse Events (AEs) in unstructured text data. Waldo is designed to assist researchers and healthcare professionals in identifying potential adverse events, such as side effects or complications, from real-world narratives like social media posts.

About the Project

Waldo leverages a fine-tuned RoBERTa model, capable of processing large volumes of unstructured text data, such as online reviews, forum discussions, and self-reported experiences, to detect potential AEs related to cannabis-derived products. The tool automates the traditionally labor-intensive process of adverse event reporting, making it easier to spot rare events that could otherwise go unnoticed.

With an impressive accuracy of 99.66%, Waldo has shown to be an invaluable asset in pharmacovigilance efforts, providing reliable and scalable AE detection.

Why Waldo?

🌟 High Accuracy: Achieves 99.66% accuracy in AE detection, outperforming many traditional methods.
📄 Comprehensive Coverage: Processes and analyzes unstructured text data from diverse sources, including social media, blogs, and medical forums.
🚀 Easy to Use: Waldo integrates seamlessly into your existing workflows, providing fast and reliable results.
🔧 Customizable: As an open-source tool, Waldo is fully customizable to meet specific research needs and can be extended or fine-tuned with additional datasets.

Key Features

High Accuracy (99.66%) in detecting adverse events from unstructured text.
Processes unstructured data from sources such as social media and medical forums.
Open-source: Free to use, customize, and contribute to.
Streamlined integration: Easily integrates into workflows, whether for research or industry.

Getting Started

To get started with Waldo, follow these steps:

Visit our GitHub Repository for detailed instructions on installation and setup.
Installation: Simple step-by-step instructions are available for installing Waldo on your local machine or server.
Usage: Learn how to run Waldo on your datasets, generate results, and analyze potential AEs.

We have provided easy-to-follow guides to get Waldo up and running for your specific use case.

Research Collaboration

Waldo is part of a larger collaborative research effort involving institutions such as Johns Hopkins University, University of California San Diego, and others. This research focuses on improving the automation of adverse event detection in the healthcare domain, particularly for cannabis-derived products.

📄 The full research paper will be available here upon publication. Stay tuned for updates.

Community and Contributions

We welcome feedback and contributions from the community! If you’d like to contribute to Waldo or have suggestions for improving the tool, please:

Open an issue in our GitHub repository.
Submit a pull request with your improvements or feature suggestions.

Together, we can make Waldo an even better tool for adverse event detection and pharmacovigilance research.

Contact Us

If you have any questions or need further assistance, feel free to reach out via our GitHub repository’s Issues page, or contact the project team directly at [email address].

Acknowledgments

This project was made possible with the support of the National Institutes of Health (NIH) / National Institute on Drug Abuse (NIDA) (UM1 DA059000) and the Burroughs Wellcome Fund for Innovations in Regulatory Science. We thank all contributors and institutions involved in making Waldo a reality.

Made with ❤️ by the Waldo Team